In　view　of　the　current　research　on　the　time　domain　characteristics　of　oil-paper　insulation　aging,　only　the　overall　state　diagnosis　is　made,　and　the　insulation　status　of　different　media　in　the　oil-paper　insulation　system　is　not　distinguished.Using　the　field　test　data　of　the　recovery　voltage　of　the　oil　paper　insulation　transformer,　the　time　domain　characteristic　quantity　was　processed　according　to　the　length　of　the　charging　time,　and　the　standard　state　index　of　insulating　oil　and　insulating　paper　was　formed.The　subjective　and　objective　weights　of　each　index　were　determined　by　scaling　expansion　method　and　coefficient　of　variation　method,　and　the　optimal　combination　of　minimum　discriminant　information　was　carried　out.Finally,　an　improved　TOPSIS　was　used　to　evaluate　the　state　of　oil-　paper　insulation　in　different　mediums.The　test　results　of　the　actual　oil　immersed　transformer　verify　that　the　method　can　realize　the　classified　assessment　of　the　insulation　status　of　the　same　transformer　oil　and　paper.At　the　same　time,　accuracy　of　the　evaluation　results　was　verified　by　comparing　the　actual　conditions　of　transformer　and　the　polymerization　degree　of　insulation　paper　and　the　water　quantity　in　oil.　©　2020,　Harbin　University　of　Science　and　Technology　Publication.　All　right　reserved.